Region Competition based Active Contour Modelling applied to Medical Image Analysis

Yanfeng Shang, Rudi Deklerck, Edgard Nyssen, Xin Yang

Research output: Chapter in Book/Report/Conference proceedingConference paper

Abstract

Active contours are among the most successful
image segmentation techniques used in a variety of
applications. Edge based object extraction is
historically the first and therefore also the most
widespread instance of the active contour model.
The disadvantage of this approach is in general that
the active contour may leak out of the ideal contour
when the edges are weak.
Region based object extracting models show more
advantages than edge based models on weak
edges. A common disadvantage of such models is
that during the execution of the iterative algorithm,
the active contour will oscillate with a high speed on
strong edges, because of the large deviation of the
features between neighboring pixels. This
sometimes leads to imprecise segmentation results.
Motivated by these problems, we studied an object
extracting model, combining edge and region
features, building further on the approach developed
in [1], called the Region Competition based Active
Contour model (RCAC).
Original languageEnglish
Title of host publicationAnnual Symposium of the IEEE/EMBS Benelux Chapter
EditorsThe Netherlands Heze
Number of pages4
Publication statusPublished - 2007
EventFinds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet - Stockholm, Sweden
Duration: 21 Sep 200925 Sep 2009

Publication series

NameAnnual Symposium of the IEEE/EMBS Benelux Chapter

Conference

ConferenceFinds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet
Country/TerritorySweden
CityStockholm
Period21/09/0925/09/09

Bibliographical note

Heze, The Netherlands

Keywords

  • Active contours
  • segmentation
  • REGION COMPETITION
  • level set

Fingerprint

Dive into the research topics of 'Region Competition based Active Contour Modelling applied to Medical Image Analysis'. Together they form a unique fingerprint.

Cite this